Abstract
This study was aimed at identifying how group knowledge is advanced and how individual learners contribute to it from the perspective of distributed cognition by using the network structure analysis. The discourse data came from a group of five university students during a course on educational theory and practices. The network analysis manifested that cognitively important turns had significantly higher betweenness centrality, and that the betweenness centrality coefficients were decreased as learning went on. With the pattern of change in betweenness centrality coefficients in the group discourse as a benchmark, each individual learner's contribution to the utterance network was discussed. Furthermore, the relationship between the learners' contributions and their final level of conceptual understanding was studied.
